Can I read Training Artificial Neural Networks by Generalized Likelihood Ratio Method: Exploring Brain-like Learning to Improve Robustness on EtoBox?

Training Artificial Neural Networks by Generalized Likelihood Ratio Method: Exploring Brain-like Learning to Improve Robustness by Xiao, Li; Peng, Yijie; Hong, Jeff; Ke, Zewu; Yang, Shuhuai is a scholarly article available to read on EtoBox.

What is Training Artificial Neural Networks by Generalized Likelihood Ratio Method: Exploring Brain-like Learning to Improve Robustness about?

In this work, we propose a generalized likelihood ratio method capable of training the artificial neural networks with some biological brain-like mechanisms,.e.g., (a) learning by the loss value, (b) learning via neurons with discontinuous activation and loss functions. The traditional back propagation method cannot train the artificial neural networks with aforementioned brain-like learning mechanisms. Numerical results show that the robustness of various artificial neural networks trained by the new method is significantly improved when the input data is affected by both the natural noises and adversarial attacks. Code is available: \url{https://github.com/LX-doctorAI/GLR_ADV} .

Author

Xiao, Li; Peng, Yijie; Hong, Jeff; Ke, Zewu; Yang, Shuhuai

Published

2019

Language

EN